Obesity and Diabetes Clinical Research Section, National Institute of Diabetes and Digestive and Kidney Diseases, National Institutes of Health, Department of Health and Human Services, 4212 North Sixteenth Street, Room 5-41, Phoenix, AZ 85016. E-mail: firstname.lastname@example.org
Objective: Increased mRNA and activity levels of 11β-hydroxysteroid dehydrogenase type 1 (11βHSD1) in human adipose tissue (AT) are associated with obesity and insulin resistance. The aim of our study was to investigate whether 11βHSD1 expression or activity in abdominal subcutaneous AT of non-diabetic subjects are associated with subsequent changes in body weight and insulin resistance [homeostasis model assessment of insulin resistance (HOMA-IR)].
Research Methods and Procedures: Prospective analyses were performed in 20 subjects (two whites and 18 Pima Indians) who had baseline measurements of 11βHSD1 mRNA and activity in whole AT (follow-up, 0.3 to 4.9 years) and in 47 Pima Indians who had baseline assessments of 11βHSD1 mRNA in isolated adipocytes (follow-up, 0.8 to 5.3 years).
Results: In whole AT, although 11βHSD1 mRNA levels showed positive associations with changes in weight and HOMA-IR, 11βHSD1 activity was associated with changes in HOMA-IR but not in body weight. 11βHSD1 mRNA levels in isolated adipocytes were not associated with follow-up changes in any of the anthropometric or metabolic variables.
Discussion: Our results indicate that increased expression of 11βHSD1 in subcutaneous abdominal AT may contribute to risk of worsening obesity and insulin resistance. This prospective relationship does not seem to be mediated by increased 11βHSD1 expression in adipocytes.
In animals and humans, glucocorticoids (GCs)1 affect energy metabolism and their elevated levels, from either endogenous or exogenous sources, can cause obesity and hyperglycemia (1,2,3). Although animal models of obesity are characterized by increased plasma levels of GC (4), in humans, morning or 24-hour plasma cortisol levels are similar or lower in obese compared with lean individuals (5,6,7). It has been proposed that tissue-specific rather than systemic increases in cortisol concentrations; thus, GC action, may influence the development of obesity (8,9). The enzyme 11β-hydroxysteroid dehydrogenase type 1 (11βHSD1) is expressed in many GC-sensitive tissues [most abundantly in adipose tissue (AT) and liver] where it amplifies GC receptor activation by converting inactive cortisone to active cortisol (8,10). Adipose 11βHSD1 activity is increased in animal models of obesity (11,12) and transgenic mice overexpressing 11βHSD1 in AT develop hyperphagia, visceral adiposity, hyperglycemia, and insulin resistance, despite normal plasma GC levels (13,14). In contrast, 11βHSD1 knockout mice exhibit increased hepatic insulin sensitivity and resistance to hyperglycemia and obesity induced by fat feeding (15).
Recently, both activity (16,17) and mRNA levels (18,19,20,21,22) of 11βHSD1 were shown to be elevated in subcutaneous AT of obese human subjects. Moreover, we have shown that in both Pima Indians and whites, 11βHSD1 activity and/or mRNA concentrations in whole AT (23) and isolated adipocytes (24) are associated with adiposity, fasting glucose, insulin levels, and insulin resistance.
However, cross-sectional relationships do not establish causality. Prospective studies are better suited to identify risk factors for the development of a disease. To evaluate the role of 11βHSD1 in the development of obesity and insulin resistance, we examined the prospective associations of baseline 11βHSD1 activity and mRNA levels in abdominal subcutaneous AT with changes in adiposity, glucose tolerance, and measures of insulin sensitivity in non-diabetic individuals. Our aims were to determine whether elevated levels of 11βHSD1 activity or mRNA in isolated adipocytes or whole AT at baseline is associated with a subsequent increase in body weight and a subsequent decline in insulin sensitivity.
Research Methods and Procedures
White and Pima Indian volunteers were recruited by newspaper advertisement or by community-based recruiters. All subjects were between 18 and 50 years of age and non-diabetic (normal glucose tolerance or impaired glucose tolerance) at baseline and follow-up according to World Health Organization criteria (25).
The subjects were non-smokers at the time of the study and, except for obesity, were healthy according to a physical examination and routine laboratory tests. For the baseline visit, all subjects were admitted for 8 to 10 days to the NIH Clinical Research Unit in Phoenix, AZ. On admission, they were placed on a weight-maintaining diet (50%, 30%, and 20% of daily calories provided as carbohydrate, fat, and protein, respectively) and abstained from strenuous exercise. Height and weight were measured with subjects dressed in light indoor clothing but without shoes. BMI was calculated as the ratio of weight (kilograms) over the square term of height (meters), and percentage body fat was assessed by DXA (DPX-L; Lunar Corp., Madison, WI) (26).
At least 3 days after admission and after a 12-hour overnight fast, subjects underwent a 2-hour 75-gram oral glucose tolerance test. Plasma glucose concentrations were determined by the glucose oxidase method (Beckman Instruments, Fullerton, CA) and plasma insulin by radioimmunoassay (Concept 4; ICN Biomedicals, Costa Mesa, CA). The homeostasis model for assessment of insulin resistance index (HOMA-IR) was calculated as previously described (27).
The individuals in this study were part of larger cohorts in whom cross-sectional relationships between 11βHSD1 expression and/or activity and anthropometric and metabolic variables at baseline were previously described (23,24). Sixty-seven subjects from the original cohorts had a follow-up visit (mean follow-up, 2.5 years; range, 0.3 to 5.3 years) for measurement of body weight and oral glucose tolerance test and were included in these analyses. These included 20 in Study 1 (two whites/18 Pima Indians, nine men/11 women) (23) and 47 in Study 2 (all Pima Indians, 29 men/18 women) (24). Follow-up measurements of percentage body fat and plasma insulin concentration were available in 42 subjects (14 in Study 1 and 28 in Study 2).
The protocol was approved by the Tribal Council of the Gila River Indian Community and by the Institutional Review Board of the National Institute of Diabetes and Digestive and Kidney Diseases. All subjects provided written informed consent before participation.
AT Biopsy Procedure
Fat biopsies were obtained during the first admission (baseline) between 8:30 am and 10 am after a 12-hour overnight fast. Subcutaneous abdominal AT was removed from the periumbilical region by percutaneous needle biopsy under local anesthesia (1% lidocaine). The adipose biopsy was placed on a sterile nylon mesh and rinsed with sterile 0.9% NaCl solution. Immediately after rinsing with saline, the tissue was frozen at −70 °C for measurement in whole AT (Study 1). For adipocyte isolation (Study 2), saline-rinsed tissue was cleaned of visible connective tissue and blood vessels in Hanks’ balanced salt solution supplemented with 5.5 mM glucose. AT was digested in Hanks’ balanced salt solution buffer containing 5.5 mM glucose, 5% fatty acid-free bovine serum albumin (Introgen/Serologicals, Norcross, GA), and 3.3 mg/mL type I collagenase (Worthington Biochemical Corp., Lakewood, NJ) for 30 minutes in a 37 °C water bath. The digestion mixture was passed through a sterile 230-μm stainless steel tissue sieve (Thermo EC, Holbrook, NY), adipocytes were allowed to float by gravity, and the supernatant containing adipocytes was collected.
RNA Extraction and cDNA Synthesis
Whole AT (Study 1)
As detailed previously, ∼500 mg of fat was homogenized in 1.5 mL of Trizol, extracted in chloroform and RNAid matrix (Anachem, Luton, Bedfordshire, United Kingdom), washed, and precipitated (23). RNA integrity was checked by agarose gel electrophoresis. Oligo dT-primed cDNA was synthesized from 0.5 μg of RNA samples. PCR amplification of glyceraldehyde 3-phosphate dehydrogenase transcript using commercial primers (Clontech, Palo Alto, CA) was carried out to confirm successful cDNA synthesis.
Isolated Adipocytes (Study 2)
As described previously, adipocyte RNA was extracted using RNeasy Mini Kit from Qiagen (Valencia, CA) (24). During the extraction, RNA was treated with RNase-free DNase (Qiagen) according to the manufacturer's instructions. One microgram of RNA from each sample was used to prepare oligo dT-primed cDNA using the Advantage RT for PCR kit (Clontech) following the manufacturer's recommendation. Successful cDNA synthesis was verified by PCR amplification of β2-microglobulin transcript using forward primer 5′-TGT CTT TCA GCA AGG ACT GGT C-3′ and reverse primer 5′-TGA TGC TGC TTA CAT GTC TCG AT-3′.
11βHSD1 mRNA Quantification (Whole AT and Isolated Adipocytes)
Quantification of 11βHSD1 mRNA concentrations was performed with real-time PCR intron spanning primer probe sets using the ABI PRISM 7700 sequence detection system (Applied Biosystems, Foster City, CA) as described previously (23,24). Real-time PCR assays for whole AT and isolated adipocytes utilized different cDNA samples for providing the standard curves. Each sample was run in duplicate, and the mean values of the duplicates were used to calculate transcript level. A standard curve for each primer probe set was generated by serial dilution of cDNA from a healthy subject done in triplicate. Reactions without template were included as negative controls in Study 2 (mature adipocytes), and reverse transcriptase negative controls were included in Study 1 (whole AT). Real-time PCR was performed as recommended by the manufacturer as follows: 50 °C, 2 minutes; 95 °C, 10 minutes; 95 °C, 15 seconds; and 60 °C, 1 minute for 40 cycles. Human cyclophilin (Applied Biosystems) was used to normalize the 11βHSD1 transcript levels.
11βHSD1 Activity Measurement (Whole AT)
Adipose 11βHSD1 activity was measured in an approximately 500-mg aliquot of the biopsies as previously described (16,17,23). Briefly, the tissue was homogenized in Krebs buffer at pH 7.4, and 750 mg/mL protein was incubated at 37 °C with 2 mM nicotinamide adenine dinucleotide phosphate and 100 nM 1,2,6,7-3H4-cortisol for 30 hours. Samples were withdrawn at 3, 6, 20, and 30 hours for separation of cortisol and cortisone by high-performance liquid chromatography (HPLC) with on-line liquid scintillation detection. The 11βHSD1 activity was measured in the dehydrogenase direction (i.e., cortisol to cortisone rather than reductase cortisone to cortisol) because reductase activity is less stable in vitro (9), and the dehydrogenase direction is preferred when the enzyme is liberated from its intracellular environment (28). When driven by excess cofactor, this activity is proportional to total protein.
Cortisol and Cortisone Concentrations in Whole AT
The infranatant from the RNA extraction protocol (see above) was used to extract steroids (20,23). Approximately 0.3 pmol/mL (<1% final tissue concentrations) of 1,2,6,7-3H4-cortisone and 1,2,6,7-3H4-cortisol (Amersham, Little Chalfont, Buckinghamshire, United Kingdom) was added to the homogenate to correct for steroid extraction and HPLC efficiency. Samples were centrifuged to remove the fat layer and extracted on a Sep-Pak (Waters C18 cartridges; Elstree, Hertfordshire, United Kingdom), further purified with hexane, reextracted with ethyl acetate, and reconstituted in mobile phase for HPLC separation of cortisone and cortisol fractions. Fractions were counted for recovery of tracer 3H-steroid (mean, 61 ± 12%) and assayed in triplicate for endogenous cortisone by radioimmunoassay (Immunovation, Southampton, United Kingdom; cross-reactivity with cortisol <0.1%) and cortisol by enzyme-linked immunosorbent assay (Salimetrics, LLC, State College, PA; cross-reactivity with cortisone, 0.31%). Final steroid concentrations are expressed per gram of wet weight of AT after adjustment for extraction efficiency.
Statistical analyses were performed using the procedures of the SAS Institute (Cary, NC). Results are presented as means ± standard deviation. To approximate a normal distribution, fasting and 2-hour insulin, HOMA-IR, 11βHSD1 mRNA and activity, tissue cortisol, and tissue cortisol-to-cortisone ratio were logarithmically transformed before statistical computations.
In the cross-sectional analyses, relationships between 11βHSD1 measures and body weight, BMI, percentage body fat, fasting and 2-hour plasma glucose, and insulin and HOMA-IR were examined by calculating Pearson correlation coefficients. Differences in anthropometric and metabolic measurements between the two visits were assessed by paired Student's t test.
Pearson correlation coefficients were calculated to estimate the correlation of baseline 11βHSD1 activity and mRNA levels, cortisol, and cortisol-to-cortisone ratio with changes in the anthropometric and metabolic variables before and after normalization for duration of follow-up. General linear models were then used to test the relationships after adjustment for age at follow-up, gender, time of follow-up, and measurement at baseline.
Study 1: 11βHSD1 in Whole Abdominal Subcutaneous AT
Anthropometric and metabolic variables of the subjects at baseline and follow-up are displayed in Table 1. On average, no significant differences between the two visits were observed in any of the anthropometric and metabolic measures.
Table 1. Baseline and follow-up clinical characteristic of subjects who had measures of 11βHSD1 in whole AT (mRNA, activity, cortisol, cortisone) or in isolated adipocytes (mRNA) at the baseline visit
Study 1: whole AT
Study 2: isolated adipocytes
11βHSD1, 11β-hydroxysteroid dehydrogenase type 1; AT, adipose tissue; Fx, fasting; HOMA-IR, homeostasis model assessment of insulin resistance.
Data are expressed as means ± standard deviation (baseline and follow-up) or range (change).
p < 0.05, variable at follow-up vs. variable at baseline.
p < 0.01, variable at follow-up vs. variable at baseline.
p < 0.001, variable at follow-up vs. variable at baseline.
In cross-sectional analyses, 11βHSD1 mRNA levels were associated with 11βHSD1 activity (r = 0.49, p = 0.03) but not with AT cortisol levels or cortisol-to-cortisone concentration ratio (p > 0.6 both). At baseline visit, 11βHSD1 activity was associated with body weight (r = 0.61, p = 0.004), BMI (r = 0.71, p = 0.0005), and waist circumference (r = 0.62, p = 0.005), whereas 11βHSD1 mRNA levels were not significantly associated with any of the anthropometric measures in this subset of the cohort. No significant correlation was found between 11βHSD1 activity or mRNA levels and fasting or 2-hour plasma glucose concentrations. At baseline, in 14 subjects who had available follow-up measurements of percentage body fat and plasma insulin concentration, 11βHSD1 activity, and mRNA levels were not associated with percentage body fat, fasting plasma insulin concentration, or HOMA-IR.
Prospectively, a positive association was observed between 11βHSD1 mRNA levels and follow-up changes in body weight (p = 0.03, Table 2). This association was present after adjustment for age, sex, weight at baseline, and time of follow-up (Figure 1). 11βHSD1 mRNA levels also showed a positive association with changes in fasting insulin levels and HOMA-IR (p = 0.03 and 0.02 respectively, Table 2). After adjustment for age, sex, HOMA-IR at baseline, change in weight, and time of follow-up, 11βHSD1 mRNA levels at baseline were significant predictors of HOMA-IR at follow-up (p = 0.02, Figure 1). No association was observed between 11βHSD1 mRNA levels and activity and changes in fasting plasma glucose concentrations (Table 2).
Table 2. Pearson correlation of 11βHSD1 and GC measures in adipocytes and whole AT with changes in body weight, waist circumference, percentage body fat, and HOMA-IR
Cross-sectionally, whole AT cortisol concentration showed a tendency to be associated with body weight (r = 0.46, p = 0.06). Cortisol concentration or cortisol-to-cortisone concentration ratio in whole AT was not associated with any of the other anthropometric and metabolic variables. Prospectively, intra-adipose cortisol concentrations and cortisol-to-cortisone ratio were not associated with changes in body weight, BMI, waist circumference, percentage body fat, fasting plasma glucose and insulin concentrations, and HOMA-IR (Table 2).
Study 2: 11βHSD1 in Isolated Abdominal Subcutaneous Adipocytes
Anthropometric and metabolic variables of the subjects at baseline and follow-up are summarized in Table 1. On average, subjects gained significant weight and waist circumference and tended to gain body fat. Fasting plasma glucose and insulin levels and HOMA-IR insulin resistance index also rose with time. The follow-up measurements of percentage body fat and insulin levels (including HOMA-IR) were available in 28 subjects only. At baseline, 11βHSD1 mRNA levels correlated with body weight (r = 0.50, p = 0.0004), BMI (r = 0.50, p = 0.001), waist circumference (r = 0.58, p < 0.0001), percentage body fat (r = 0.37, p = 0.05), fat-free mass (r = 0.47, p = 0.01), fat mass (r = 0.49, p = 0.008), fasting insulin concentration (r = 0.49, p = 0.008), and HOMA-IR (r = 0.43, p = 0.02). Prospectively, there were no associations between 11βHSD1 mRNA levels in isolated adipocytes and follow-up changes in any of the anthropometric and metabolic variables (Table 2).
We investigated prospective relationships between 11βHSD1 activity and mRNA levels in abdominal subcutaneous AT and subsequent changes in body weight and composition and insulin resistance. Higher baseline 11βHSD1 mRNA levels in whole AT were associated with larger adverse changes in body weight and insulin resistance (assessed by HOMA-IR). In contrast, 11βHSD1 expression in isolated adipocytes showed no relationship with future changes in body weight and insulin resistance.
11βHSD1 is expressed widely throughout the body, and its overexpression within AT has been proposed to be important in the development of obesity and insulin resistance (9). Although significant associations between 11βHSD1 mRNA or activity in subcutaneous AT and obesity or insulin resistance have been reported in several cross-sectional studies (16,17,18,19,20,22,23), this finding is not universal (29). Importantly, neither prospective nor longitudinal investigations have been performed to support a causal role of 11βHSD1 in the etiology of obesity and/or insulin resistance in humans. Association studies of microsatellite markers and single nucleotide polymorphisms in the 11βHSD1 gene have not supported a significant relationship of these variations with obesity, albeit that they do predict type 2 diabetes and hypertension (24,30).
Results from this first prospective human study show that, although whole subcutaneous abdominal AT 11βHSD1 mRNA and activity predicts subsequent weight gain and/or insulin resistance, the expression levels of 11βHSD1 in the isolated abdominal subcutaneous adipocytes do not predict progressive obesity. The contrasting prediction from 11βHSD1 mRNA levels in whole AT compared with isolated adipocytes may reflect involvement of 11βHSD1 expressed by cells of AT stroma such as adipocyte precursors (preadipocytes) and macrophages (29,31). Cortisol is crucial in promoting differentiation of preadipocytes to adipocytes (32). Increased expression of 11βHSD1 in preadipocytes may stimulate conversion of cortisone to cortisol, thus stimulating adipogenesis resulting in increased propensity to weight gain.
It must be noted, however, that there were significant differences in group characteristics. Unlike the subjects in Study 2, as a group, the subjects in Study 1 did not show significant changes in any of the anthropometric and metabolic outcomes between the baseline and follow-up visits. Moreover, the group in Study 1 was smaller, which may increase the risk of false-positive results. The isolation of adipocytes itself might also modulate 11βHSD1 mRNA levels in adipocytes as has been shown for several proadipogenic and inflammatory genes (33). Adipocyte enlargement is strongly associated with insulin resistance and predicts type 2 diabetes (35). The disruption of large adipocytes during the isolation procedures (34) and, thereby, a lack of them in Study 2, may also influence the relationship of 11βHSD1 mRNA levels with prospective changes in insulin resistance.
We did not find any relationship between intra-adipose cortisol and cortisol-to-cortisone levels and follow-up changes in anthropometric and metabolic outcomes. Unlike animal models (13,14), intra-adipose levels of cortisol in cross-sectional human studies are unrelated to 11βHSD1 activity or mRNA levels (20,23), and a similar result was observed in cross-sectional analyses of our subjects at baseline. There are considerable limitations of intra-adipose cortisol-to-cortisone measurements in our study that have been more extensively discussed in our previous articles (20,23). First, the extraction technique we used for the relatively small aspiration biopsy samples was novel. Second, it has been shown that subjects with metabolic syndrome may respond to invasive procedures by exaggerated cortisol release (36). The information about the effect of biopsy stress on cortisol secretion is lacking in our study because concomitant plasma samples were not obtained. Finally, the measurements were performed in the decreasing phase of the diurnal cycle while local production could be most important at the nadir of adrenal secretion of GCs (9).
In conclusion, our prospective analyses show that 11βHSD1 expression and activity in whole subcutaneous abdominal AT, but not the adipocyte fraction, predicted later anthropometric and metabolic outcomes over a follow-up period of an average 2.5 years. These results suggest that increased 11βHSD1 expression in abdominal subcutaneous adipocytes may represent a marker rather than a determinant of obesity and insulin resistance, whereas 11βHSD1 activity in other AT cells may have a causative role in promoting obesity and its metabolic consequences.
We gratefully acknowledge the help of Thomas Brookshire and Pamela Daychild in subject recruitment and the nursing and dietary staffs of the NIH Metabolic Unit for the care of the volunteers. We are grateful to the members and leaders of the Gila River Indian Community for their continuing cooperation in our studies. This work was funded by the intramural research program of the National Institute of Diabetes and Digestive and Kidney Diseases/NIH/Department of Health and Human Services and the British Heart Foundation (to B.R.W. and D.J.W.).
Nonstandard abbreviations: GC, glucocorticoid; 11βHSD1, 11β-hydroxysteroid dehydrogenase type 1; AT, adipose tissue; HOMA-IR, homeostasis model for assessment of insulin resistance index; PCR, polymerase chain reaction; HPLC, high-performance liquid chromatography.
The costs of publication of this article were defrayed, in part, by the payment of page charges. This article must, therefore, be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.